1. Field of the Invention
In general, the present invention relates to systems and methods that are used to identify a person using a visible biometric pattern that is unique to that person. More particularly, the present invention relates to systems and methods that illuminate a person and capture an image of at least some portion of that person's biometric features so that the image can be analyzed and identified.
2. Prior Art Description
Every person has unique biometric identifiers. For example, every person has a unique fingerprint and a unique DNA sequence. However, such biometric identifiers cannot be obtained simply by imaging a person as they pass a camera. However, other biometric identifiers can be readily perceived from a camera image. Those identifiers include face recognition identifiers and iris pattern identifiers.
There are many existing face recognition algorithms that can identify a person from an image of their face. Such algorithms typically measure proportions of the face, such as the distance between the eyes, the position of the nose and the position of the mouth. Since the proportions of most faces differ slightly, individuals can be identified by measuring such proportions from clear captured images. Prior art face identification systems are exemplified by U.S. Pat. No. 7,203,346 to Kim, entitled Face Recognition Method And Apparatus Using Component Based Face description; and U.S. Pat. No. 6,301,370 to Steffens, entitled Face Recognition From Video Images.
A more accurate identification methodology can identify a person from that person's iris. The iris is the colored portion of the eye that surrounds the pupil. Upon close inspection, it can be seen that the iris is not a single monochromatic color. Rather, the iris has a complex pattern of overlapping colors, lines, and speckles that extend throughout the iris. The iris pattern of any one individual is as unique to that individual as his/her fingerprints. Consequently, it is highly improbable that any two individuals in the world would share the same iris pattern.
Since the iris pattern of an individual is such a good biometric identifier, there have been many prior art systems developed to image the iris of an individual and use that image for identification purposes. Such prior art systems are exemplified by U.S. Pat. No. 5,291,560 to Daugman, entitled, Biometric Personal Identification System Based On Iris Analysis; U.S. Pat. No. 7,277,561 to Shin, entitled Iris Identification; and U.S. Pat. No. 7,796,784 to Kondo, entitled Personal Authentication Method For Certificating Iris. In all such prior art systems, a clear image of the iris is required. That can present a problem, in that obtaining a clear image of a person's face and/or iris in the real world is very difficult.
The largest problem associated with obtaining a clear image of person's face and/or the iris is one of proper illumination. In order to obtain a clear and reliable image, the features of the person must be illuminated with light that is brighter than that of the ambient background light, else reflections and/or shade regions can contaminate the pattern being imaged. However, background full spectrum sunlight can sometimes be as bright as 100 mW/cm2. In order to overcome this level of background light, a face must be illuminated with such an intense flash that the person would be momentarily blinded or the person would experience at least some physical discomfort from the light's intensity. Furthermore, repeatedly illuminating the same person with such an intense light may result in some retinal damage.
One solution tried to solve the problem of illumination is to illuminate the person with infrared light instead of white light. The human eye is less sensitive to infrared light than other shorter wavelengths of visible light. In the prior art, infrared light is typically created with infrared LEDs, due to the commercial availability of these LEDs. However, a very large matrix of infrared LEDs would have to be used in order to surpass the infrared light contained in background sunlight. Furthermore, although the eye is far less sensitive to such infrared light, the use of infrared light greatly reduces the contrasts of the pattern being imaged. This is due to the fact that light produced by commercial LEDs is generally very narrowband, if not monochromatic. The small bandwidth of wavelengths being produced makes it more difficult to detect finely detailed patterns in the image, which the uniqueness being sought. Since most images used to identify people contain subtle details, a lot of pattern information is lost. Therefore, although the iris may be illuminated by LED light, the image obtained lacks much of the contrast detail needed for many pattern identification algorithms to function to their full potential.
Another problem associated with aggressive illumination is one of producing specularities. Specularities are the areas of the eye or eyeglasses that reflect the illuminating light back into the camera and cause an image saturation and obscuration of many details being imaged. The reflected light appears as a white, or saturated area in the captured image, wherein no iris pattern information can be obtained. Specularities occur with the naked eye from the cornea. However, these small corneal specularities rarely obscure the entire image. However, if the person being imaged is wearing eyeglasses, then the eyeglasses may act as mirrors, depending upon the angle of the eyeglasses. This can produce larger specularities that make a significant portion of captured image unreadable, which impacts overall system performance.
The obvious solution to the above-identified problems is to eliminate background illumination and specularities by making a person remove his/her glasses and place his/her face directly in front of a scanner. In this manner, a person's face occupies most of the scanned image and blocks out background light. This close-proximity scan eliminates most background lighting problems and most specularities. Although close proximity scanners may be appropriate for certain applications, such as access through high security doors, such scanners have little practical use in public areas, such as airports, train stations or the like. Nor are close proximity scanners practical for outdoor environments where bright sunlight is prevalent. Likewise, close proximity scanners have no applications in passive monitoring of crowds, wherein people do not stop to be scanned but, rather, are automatically scanned as they pass a certain point.
In order for a scanner to passively monitor a crowd, the imaging camera must be focused at some point within the crowd or an auto-focusing system must be utilized. In order to achieve higher device reliability, smaller form factors, and reduced costs, the imaging lens system must have an illumination system sufficient for the depth of field or static capture zone depth. The high level of illumination is compatible with the higher lens F# and produces a deeper static capture zone without an auto-focusing system. Different static capture zone designs are achieved by using different lenses designed to achieve various capture zone distances from several inches to beyond ten meters. At capture zone distances beyond ten meters, the greater depth of field from the high F# reduces the requirements of the auto-focusing system by providing a more favorable depth of field. It has proven difficult in the prior art to provide sufficient illumination throughout such an extended range without making the illumination flash either highly obvious or potentially harmful.
A need therefore exists for a system and method, whereby a usable image can be obtained regardless of worst-case ambient lighting conditions. A need also exists for a system and method of obtaining a clear image without requiring a person to remove his/her eyeglasses. A further need exists for a system and method that can reliably image a person's biometric features from at distances few inches to ten meters with adequate illumination, therein eliminating the need for a person to stand in close proximity to a scanner. Lastly, a need exists for a system that illuminates a person in a manner that is not obvious to the person being scanned, optically annoying, and/or potentially harmful. These needs are met by the present invention as described and claimed below.